High-Availability Allows Business Continuity, Says Dietmar Maurer, Proxmox CTO


Proxmox Server Solutions GmbH — based in Vienna, Austria — offers enterprise server virtualization solutions, including the open source project Proxmox Virtual Environment (VE), which combines container-based virtualization and KVM/QEMU on one web-based management interface. The company was founded in 2005 by brothers Martin and Dietmar Maurer. In 2014, the company joined the Linux Foundation to deepen its commitment to virtualization technologies such as KVM.

In this exclusive interview, Dietmar Maurer, CTO of Proxmox, talks about how virtualization is driving the modern IT infrastructure and how high availability (HA) directly affects business operations.

Dietmar Maurer, CTO of Proxmox

Linux.com: Can you tell us how virtualization helps businesses get the most out of their IT infrastructure in this new era of software-defined everything?

Dietmar Maurer: Software-defined storage (SDS) via open source technologies is the final step to operate all services without any vendor lock. Recently, most enterprises used traditional storage like iSCSI or NFS from tier one vendors for their virtualization cluster setups. Nowadays, many businesses have started moving to software-defined storage, like Ceph for example — just to name the most important one being used also with our virtualization platform Proxmox VE.

Together with the wide availability of enterprise class NVMe SSD and 10 or 40Gbit networks, high-performance storage clusters are already in place – the upcoming next generation of enterprise SSD storage (Intel 3D Xpoint NVMe SSD) will boost SDS again to an even greater level. The great thing for all open source users is that they can benefit immediately from such hardware innovation.

Virtualization has become the norm in modern IT infrastructure, but some businesses are just starting to adopt virtualization platforms. What are the key areas they should consider when moving to virtualized environments?

If possible, I would recommend to always use open technologies. Traditional software vendors always will tell you that their closed solution is ten times faster. But, you should take your time and evaluate closely several offerings with your own test lab. You will then be able to see the difference it makes in reality, and in numbers. You should always try to avoid bottlenecks in your IO stack and also make sure that your storage is fast enough (use high end SSD only!). As long as you choose a scalable and expandable storage solution, you are prepared for the future needs of an expanding business.

What challenges do businesses face in a virtualized environment when it comes to provisioning/availability of applications, services, databases, networks, storage, etc.?

Everybody needs highly available services, nobody likes downtime – this is true for the one-person company as well as for big enterprises. Building HA setups with a fully redundant network and storage is not so hard, and it’s possible for any admin. A common mistake, in my experience, is that people choose cheap hardware for HA nodes. This should be always avoided as it leads to a bunch of problems afterward. Only premium class server hardware should be used here.

We know that in the traditional model, provisioning of IT resources can take days, weeks, and even months. How do VMs and containers help solve that problem and what other benefits do they bring?

Almost all servers run on Linux nowadays. Choosing a Linux-based virtualization, Linux-based network, and Linux-based software storage seems obvious. As a result of the common Linux basis of your data center software, automation on almost all places is much easier than by mixing different technologies. By using separate containers or virtual machines for the needed services, the management and operation can be secured and optimized. The ability to live-migrate virtual machines from one hardware to another keeps all your services alive, regardless of hardware replacement or maintenance tasks in your data center.

What kind of businesses need high availability of virtualized environments?

Today, everyone dependent on IT needs high availability, so actually all businesses who have IT need HA to minimize server downtime. Imagine, for example, a business with an online shop selling a product X. The server breaks down and is offline for a couple of hours or, worst case, for even a day. A client visiting the shop at this time sees that it’s not available. He will buy somewhere else, and it’s doubtful he’ll ever come back again.

Let me explain how high availability is used in Proxmox VE: If a virtual machine or container (VM/CT) is configured as HA and the physical host fails, the VM or CT is automatically restarted on one of the remaining cluster nodes. In Proxmox VE, system administrators are able to configure complex HA cluster settings intuitively via the web GUI, which is why Proxmox VE makes high availability easily accessible to the masses. By integrating a software watchdog, the external fencing devices become dispensable in basic configurations.

So, to summarize, highly available virtualized environments are for all businesses who need to meet customer expectations, provide stability and reliability, and who want to grow and expand. Zero disruptions or inconveniences, minimal downtime – all this will show reliability to your customers or users. Companies offering services that should be delivered continuously and reliably — such as websites or web services (like online stores) — are just one of many examples. A reliable network helps you focus on growth instead of fixing issues or interruptions of network.

How do you ensure integrity, security, and redundancy of critical data?

This is a quite general question, and therefore I can give just a general answer. At Proxmox, we integrate the best available open source storage technologies into Proxmox VE. On the storage level, this is, for example Ceph, ZFS on Linux, DRBD, GlusterFS, and others.

Configuration files are stored in our Proxmox Cluster file system (pmxcfs). pmxcfs is a database-driven file system for storing configuration files, replicated in real time on all nodes using corosync. We use this to store all PVE-related configuration files. Although the file system stores all data inside a persistent database on disk, a copy of the data resides in RAM.

How does HA directly affect business operations?

HA allows business continuity. If your services are offline, you cannot make money. It’s that simple.

What kind of HA solutions for virtualized platforms are available for such customers?

They can use for example the Proxmox VE HA resource manager for multi-node high availability clusters. It monitors all virtual machines and containers on the whole cluster and gets automatically into action if one of them fails. The HA Manager works out of the box, and with watchdog-based fencing simplifies deployments dramatically. The entire HA settings can be configured via the integrated web GUI. It’s open source, it’s easy to implement and administer, so I think there is no excuse to not have a HA solution setup even if you are a small company.

Can you give us a brief overview of Proxmox, the company?

Proxmox Server Solutions is the company behind the open source project Proxmox Virtual Environment (VE). My brother Martin Maurer (CEO) and I founded the company in 2005 when we had developed a new product, the anti-spam and anti-virus filter Proxmox Mail Gateway, which we started selling with the new company and via our partners. In 2008, we released Proxmox Virtual Environment (VE), a server virtualization solution combining container-based virtualization and KVM/QEMU on one web-based management interface.

We had been using KVM and containers (OpenVZ back then) internally for development purposes and needed a GUI for it. That’s actually how Proxmox VE was born, and it was — and still is — a unique and really cool product; that’s why we made it available to public. Today, Proxmox also offers services like support subscriptions and trainings to our 6500+ customers in over 140 countries. We have a core team located in the beautiful city of Vienna, Austria and many contributors, testers, friends, and developers, which can be spotted all over the planet.

As a Linux Foundation member, how do you engage with the Linux and open source community?

We jumped into Debian GNU/Linux almost 20 years ago. During this time, we contributed and worked on many places and projects as we love to communicate and share ideas. Cooperating with many great open source communities allows us to access all the latest and greatest technologies without delay, and make them also available for our users.

Our main project, Proxmox VE, is licensed under the free software license AGPLv3, and there are no hidden features or commercially licensed add-ons. Besides the open source license, the development process is also totally open, and we warmly invite people to contribute and share their ideas.